Alcoholism is a chronic relapsing illness in which heavy alcohol use results in significant neuroadaptations in brain stress, reward and emotion circuits associated with high stress, alcohol craving and increased risk of relapse. In the previous funding cycle, we identified significant biobehavioral and neural predictors of future alcohol relapse. Hyperactivity of ventromedial prefrontal cortex (VmPFC) during neutral-relaxed states emerged as a primary predictor, meeting sensitivity and specificity criteria for biomarker development, while secondary predictors including alterations of basal morning cortisol and cortisol/ACTH ratio (a measure of adrenal sensitivity), stress-induced blunted Cort/ACTH ratio, stress- and cue-induced craving, smaller gray matter volume (GMV) in medial frontal and parietal-occipital regions, and blunted brain VmPFC and ventral striatal (VS) response during stress and cue-induced craving also predicted future relapse. Thus, extending previous findings, and applying a novel multi-method translational approach, we hypothesize that these alcohol-related neuroendocrine and neural changes are observable in acute and protracted abstinence and can accurately classify future relapse and treatment outcome in separate patient samples, thereby validating them as biomarkers of relapse with potential clinical utility as prognostic markers for identifying and treating those most susceptible to relapse in recovery. In the next 5-year funding period, we propose to assess these primary and secondary predictors in a prospective cohort of 100 treatment engaged abstinent alcohol dependent (AD) individuals and 50 demographic and handedness matched control drinkers (HC), and in an additional 60 AD patients enrolled in a currently ongoing RCT of Prazosin/Placebo (PZ/PL) to determine their predictive power in accurate classification of alcohol relapse and treatment outcomes. The following specific aims are proposed: (1) To examine chronic alcohol-related changes in primary and secondary predictors in AD patients during acute and protracted alcohol abstinence, and in AD patients relative to social drinkers. (2) To examine whether neuroadaptations in primary and secondary predictors during acute and protracted abstinence are each predictive of future relapse, and determine receiver operating characteristics (ROC) for relapse classification accuracy and determine cut-off values for biomarker development and testing. (3) To evaluate whether the primary and secondary predictors improve with PZ/PL treatment and predict alcohol treatment outcomes in the RCT. (4). To assess the effects of demographic (sex, age, race) and clinical (executive function, stress/trauma) variables on primary and secondary predictors and on relapse risk. Exploratory Aim: To assess additional neurophysiological assessments such as heart rate variability, plasma immune responses, diffusion tensor imaging (DTI) of white matter integrity and resting state functional connectivity (rsFC) for changes during abstinence and for prediction of relapse. Successful completion of the proposed aims has the potential to identify the first set of clinical biomarkers of alcohol relapse that may be further developed for clinical use and be targeted for treatment development to significantly improve recovery outcomes in alcoholism.